Log Management¶
Omnia maintains logs across playbook executions, container operations, and cluster activities. These logs are essential for monitoring, debugging, and troubleshooting deployments across the OIM, Kubernetes, and Slurm environments.
Warning
Do not delete log files or the directories they reside in.
Tip
To generate log files for a specific playbook, use the cd command to move into the playbook directory before running it. For example, run cd local_repo before executing the playbook to get dedicated local repo logs. If the directory is not changed, all playbook execution logs are consolidated under /opt/omnia/log/core/playbooks.
Log locations¶
Playbook logs¶
All Ansible playbook execution logs are written to the OIM host filesystem under /opt/omnia/log/core/playbooks/:
| Log File | Purpose | Playbook |
|---|---|---|
discovery.log |
Discovery logs | discovery/discovery.yml |
local_repo.log |
Local repository logs | local_repo/local_repo.yml |
prepare_oim.log |
Prepare OIM logs | prepare_oim/prepare_oim.yml |
provision.log |
Provision logs | provision/provision.yml |
telemetry.log |
Telemetry logs | telemetry/telemetry.yml |
utils.log |
Utility logs | utils/*.yml |
credential_utility.log |
Credential utility logs | credential_utility/get_config_credentials.yml |
validation_omnia_project_default.log |
Input validation report logs | |
input_validation.log |
Input validation playbook logs | input_validation/validate_config.yml |
gitlab_build_stream.log |
GitLab BuildStreaM logs | |
build_image_<arch>.log |
Build image logs | build_image_<arch>/build_image_<arch>.yml |
Core and container logs¶
| Location | Purpose |
|---|---|
/opt/omnia/log/openchami/*log |
OpenCHAMI playbook logs |
/opt/omnia/log/pulp/*log |
Pulp container logs |
/opt/omnia/log/local_repo/*log |
Local repo logs |
/opt/omnia/log/core/container/*log |
Core container logs |
/opt/omnia/log/build_stream/ |
BuildStreaM pipeline logs |
Note
BuildStreaM and GitLab log paths are available inside the BuildStreaM container.
Slurm logs¶
On Slurm cluster nodes, logs are stored in standard Slurm log directories:
| Path | Description |
|---|---|
/var/log/slurm/slurmctld.log |
Slurm controller daemon log (on the control node). |
/var/log/slurm/slurmd.log |
Slurm compute daemon log (on each compute node). |
/var/log/slurm/slurmdbd.log |
Slurm database daemon log (job accounting). |
Viewing Podman container logs on OIM¶
-
List all containers running on the OIM:
Run on: OIM hostpodman ps -aExpected outputCONTAINER ID IMAGE COMMAND CREATED STATUS PORTS NAMES 222d8d96554b localhost/omnia_auth:1.0 /bin/sh -c mkdi… 2 days ago Up 2 days 0.0.0.0:389->389/tcp, 0.0.0.0:636->636/tcp omnia_auth 0640a9adfce3 localhost/omnia_core:2.2 2 days ago Up 2 days 2222/tcp omnia_core 42515184e9ba docker.io/pgsty/minio:RELEASE.2026-04-17T00-00-00Z server /data –co… 2 days ago Up 2 days 0.0.0.0:9000-9001->9000-9001/tcp minio-server 5bbce8efdc27 docker.io/pulp/pulp:3.80 /init 2 days ago Up 2 days 0.0.0.0:2225->2225/tcp, 80/tcp pulp d0b76ac340c7 docker.io/library/postgres:16 postgres 2 days ago Up 2 days 5432/tcp omnia_postgres 0608aa923b7c docker.io/dellhpcomniaaisolution/omnia_build_stream:1.1 2 days ago Up 2 days omnia_build_stream 251709e17ec2 docker.io/library/registry:3.1.0 /etc/distribution… 2 days ago Up 2 days 0.0.0.0:5000->5000/tcp registry 992a9ca91d5d docker.io/library/postgres:11.5-alpine postgres 2 days ago Up 2 days 5432/tcp postgres b731b88c87ff ghcr.io/openchami/local-ca:v0.2.6 /step-ca.sh 2 days ago Up 2 days 9000/tcp step-ca ec2a4422424d docker.io/oryd/hydra:v2.3 serve -c /etc/con… 2 days ago Up 2 days hydra 0751a15ca614 ghcr.io/openchami/opaal:v0.3.12 /opaal/opaal serv… 2 days ago Up 2 days opaal-idp 72e566933565 ghcr.io/openchami/smd:v2.19.3 /smd 2 days ago Up 2 days 27779/tcp smd e95d50f8da14 ghcr.io/openchami/opaal:v0.3.12 /opaal/opaal logi… 2 days ago Up 2 days opaal 3fac48412fe9 ghcr.io/openchami/bss:v1.32.2 /bin/sh -c /usr/l… 2 days ago Up 2 days 27778/tcp bss 81f488e6b5cb ghcr.io/openchami/cloud-init:v1.3.0 /usr/local/bin/cl… 2 days ago Up 2 days cloud-init-server 7e970ad459ca cgr.dev/chainguard/haproxy:latest haproxy -f /usr/l… 2 days ago Up 2 days 0.0.0.0:8081->80/tcp, 0.0.0.0:8443->443/tcp haproxy 5f5028bab1cd ghcr.io/openchami/coresmd:v0.4.3 /coredhcp 2 days ago Up 2 days coresmd-coredhcp 354f3305a7b8 ghcr.io/openchami/coresmd:v0.4.3 /coredns 2 days ago Up 2 days coresmd-corednsNote
Container IDs and image versions will vary on every system.
-
View logs from a specific container:
Run on: OIM hostpodman logs <container_name> -
If the container is managed as a systemd service, view logs with:
Run on: OIM hostjournalctl -xeu <container_name>
Viewing Kubernetes pod logs¶
-
List all namespaces and pods:
Run on: Kubernetes control planekubectl get pods -A -
Get the list of containers in a pod:
Run on: Kubernetes control planekubectl get pods <pod_name> -o jsonpath='{.spec.containers[*].name}' -
View logs for a specific container:
Run on: Kubernetes control planekubectl logs <pod_name> -n <namespace> -c <container_name>
Cluster log collection¶
Omnia provides a log collection playbook for gathering cluster logs from Kubernetes and Slurm nodes for debugging and support.
ssh omnia_core
cd /omnia/log_collector
ansible-playbook collect.yml
Collection modes¶
-
Full mode (default): Collects all logs from target nodes.
Run on: omnia_core containeransible-playbook collect.yml -
Curated support mode: Excludes temporary and stale log files.
Run on: omnia_core containeransible-playbook collect.yml --tags curated_support
Output artifacts¶
| Artifact | Description |
|---|---|
| Workspace | /opt/omnia/collector_logs |
| Bundle | omnia_logs_<YYYYMMDD-HHMMSS>.tar.gz |
| Metadata | metadata.json (included in bundle) |
| Checksum | .sha256 file for integrity verification |
Note
- PXE mapping file must exist at
/opt/omnia/input/project_default/. - Nodes must be reachable from the OIM.
Warning
Logs are stored in /var. Ensure sufficient space is allocated to the /var partition.
Info
- General Troubleshooting -- Uses logs as a primary diagnostic tool.
- Best Practices Checklist -- Storage and maintenance best practices.